UNAIDS statistics estimate that approximately 800,000 infants are infected with HIV annually, with 90% of infections occurring in resource-limited settings. Virtually all HIV infections in infants are attributed to mother-to-child transmission (MTCT), which can occur during the periods of pregnancy, birth, or breast- feeding. In the absence of treatment, approximately 25% of infants born to HIV-positive mothers will be infected with HIV. In resource-rich settings, treatment with highly-active anti-retroviral therapies (HAART) has reduced the transmission rate to under 2%; However, determining the optimal time to test and providing early, timely treatment remain areas of concern. The gold standard tests for diagnosis of HIV infection in infants are HIV-1 DNA and RNA assays. Previous work has shown that the time to positive detection of infection varies significantly among infected infants. Current knowledge is based on small studies conducted primarily in the U.S./Europe and in populations who are, at most, exposed to monotherapy regimens. Previous work on the timing of MTCT suffers from several limitations, including limited sample size and the use of ad hoc strategies. We propose to characterize assays to detect HIV infection in infants and to investigate the timing of MTCT, based on a combined analysis of data from several cohorts of HIV-exposed infants. Central to our proposal is the formation of an interdisciplinary collaboration of investigators from several cohorts including approximately 5,800 HIV-positive formula-fed and breastfed infants born to HIV-positive mothers representing populations in the U.S., Europe, Africa, and Asia. The resulting large database will enable statistical analyses to evaluate the performance of DNA/RNA assays in diverse settings, including in populations exposed to HAART and in settings where non-B subtype infections dominate. We will apply novel statistical models for the estimation of timing of MTCT that overcome disadvantages inherent in previously employed ad hoc strategies. Specifically, our proposal addresses the following: Aim 1 - Characterization and comparison of the distribution of time to positive signal of HIV-1 DNA and RNA assays in HIV-1 infected, non-breastfed infants; Aim 2a - Estimation of timing of MTCT of HIV in HIV-infected, non- breastfed infants; and Aim 2b - Estimation of the timing of MTCT of HIV in HIV-infected, breastfed infants. In each aim, special emphasis will be placed on evaluating maternal/infant factors, including type of maternal/infant antiretroviral treatment, HIV subtype, and mode of delivery; mother's age, HIV-1 viral load, and CD4 cell count; and infant's CD% and birth weight. The results gathered from Aim 2b, will serve as a foundation for a larger meta-analysis of risks and factors involved in MTCT of HIV in populations where breastfeeding is the norm. IMPACT: A better understanding of the characteristics of diagnostic assays and the timing of MTCT will guide strategies to optimize HIV diagnostic testing in infants in various populations, and will inform the best approaches for scheduling prophylaxis, particularly when resources are limited.